Building a More Predictive Protein Force Field: A Systematic and Reproducible Route to AMBER-FB15 - PubMed
- ️Sun Jan 01 2017
Building a More Predictive Protein Force Field: A Systematic and Reproducible Route to AMBER-FB15
Lee-Ping Wang et al. J Phys Chem B. 2017.
Abstract
The increasing availability of high-quality experimental data and first-principles calculations creates opportunities for developing more accurate empirical force fields for simulation of proteins. We developed the AMBER-FB15 protein force field by building a high-quality quantum chemical data set consisting of comprehensive potential energy scans and employing the ForceBalance software package for parameter optimization. The optimized potential surface allows for more significant thermodynamic fluctuations away from local minima. In validation studies where simulation results are compared to experimental measurements, AMBER-FB15 in combination with the updated TIP3P-FB water model predicts equilibrium properties with equivalent accuracy, and temperature dependent properties with significantly improved accuracy, in comparison with published models. We also discuss the effect of changing the protein force field and water model on the simulation results.
Figures

Plot of the potential energy in alanine dipeptide calculated for energy-minimized structures at the MP2/aug-cc-pVTZ level with the (φ,ψ) dihedral angles constrained. Color indicates the relative potential energy with respect to the minimum.

MM vs. QM potential energies for MM-optimized geometries of threonine dipeptide. Each data point corresponds to a local energy minimum predicted by the force field. The Cycle 1 parameters were fitted to QM data from the torsion scans only. The QM data points at the local minima of Cycle 1 are added to the optimization of the Cycle 2 parameters. Cycle 3 is the final parameter set. The spurious MM energy minima (points far below the diagonal line) are eliminated in later cycles.

Time series of RMSD to the experimental crystal structure for three proteins and four simulations. The right panels show Gaussian kernel density estimates of the RMSD values. The diamond markers indicate the RMSD to the crystal structure using the Cartesian-averaged protein backbone conformations over the whole simulation.

Lipari-Szabo S2 order parameters and error residuals compared to experimental NMR measurements. The root-mean-squared error (RMSE) and mean signed error (MSE) of the simulated observables with respect to experiment are given in the legends. The background of the error residual plots are colored according to secondary structure as determined by DSSP analysis (white, helix; light gray, coil; dark gray, strand).

Scatter plots of experimental vs. calculated NMR three-bond scalar couplings. Two proteins are shown (left: bacteriophage lysozyme, PDB ID 1AM7, right: GB3, PDB ID 1IGD) and three models (top, AMBER99SB-ildn/TIP3P; middle, AMBER99SB-nmr/TIP3P; bottom, AMBER-FB15/TIP3P-FB from this work.) Symbols represent the atom pair involved in the coupling, and colors represent the position of the residue in the protein sequence.

Temperature dependence of secondary structure for two small peptides as a function of temperature and several force field / water model combinations. The performance of the AMBER-FB15 / TIP3P-FB model combination is the dark blue trace in the middle row. Left column: The helical fraction of Ac-(AAQAA)3-NH2. Right column: The fraction folded of CLN025. Top row: Comparison of multiple protein force fields using TIP3P water model. Middle row: Same comparison using TIP3P-FB water model. Bottom row: Comparison of four water models using AMBER-FB15 protein force field.

Correlation between average protein-water interaction energy and fraction of secondary structure. Left: AAQAA simulated with A99SB-V (left). Right: CLN025 simulated with AMBER-FB15 (right). Each plot contains four simulations with four water models. Error bars represent one standard error.
Similar articles
-
Stoppelman JP, Ng TT, Nerenberg PS, Wang LP. Stoppelman JP, et al. J Phys Chem B. 2021 Nov 4;125(43):11927-11942. doi: 10.1021/acs.jpcb.1c07547. Epub 2021 Oct 20. J Phys Chem B. 2021. PMID: 34668708 Free PMC article.
-
Gopal SM, Wingbermühle S, Schnatwinkel J, Juber S, Herrmann C, Schäfer LV. Gopal SM, et al. J Phys Chem B. 2021 Jan 14;125(1):24-35. doi: 10.1021/acs.jpcb.0c08702. Epub 2020 Dec 31. J Phys Chem B. 2021. PMID: 33382616
-
Improved side-chain torsion potentials for the Amber ff99SB protein force field.
Lindorff-Larsen K, Piana S, Palmo K, Maragakis P, Klepeis JL, Dror RO, Shaw DE. Lindorff-Larsen K, et al. Proteins. 2010 Jun;78(8):1950-8. doi: 10.1002/prot.22711. Proteins. 2010. PMID: 20408171 Free PMC article.
-
Dauber-Osguthorpe P, Hagler AT. Dauber-Osguthorpe P, et al. J Comput Aided Mol Des. 2019 Feb;33(2):133-203. doi: 10.1007/s10822-018-0111-4. Epub 2018 Nov 30. J Comput Aided Mol Des. 2019. PMID: 30506158 Review.
-
Thermodynamic properties for applications in chemical industry via classical force fields.
Guevara-Carrion G, Hasse H, Vrabec J. Guevara-Carrion G, et al. Top Curr Chem. 2012;307:201-49. doi: 10.1007/128_2011_164. Top Curr Chem. 2012. PMID: 21678137 Review.
Cited by
-
System-Specific Parameter Optimization for Nonpolarizable and Polarizable Force Fields.
Hu X, Amin KS, Schneider M, Lim C, Salahub D, Baldauf C. Hu X, et al. J Chem Theory Comput. 2024 Feb 13;20(3):1448-1464. doi: 10.1021/acs.jctc.3c01141. Epub 2024 Jan 27. J Chem Theory Comput. 2024. PMID: 38279917 Free PMC article.
-
Zou J, Simmerling C, Raleigh DP. Zou J, et al. J Phys Chem B. 2019 Dec 12;123(49):10394-10402. doi: 10.1021/acs.jpcb.9b08323. Epub 2019 Dec 3. J Phys Chem B. 2019. PMID: 31702919 Free PMC article.
-
Hudson PS, Han K, Woodcock HL, Brooks BR. Hudson PS, et al. J Comput Aided Mol Des. 2018 Oct;32(10):983-999. doi: 10.1007/s10822-018-0165-3. Epub 2018 Oct 1. J Comput Aided Mol Des. 2018. PMID: 30276502 Free PMC article.
-
Giese TJ, York DM. Giese TJ, et al. J Chem Theory Comput. 2019 Oct 8;15(10):5543-5562. doi: 10.1021/acs.jctc.9b00401. Epub 2019 Sep 17. J Chem Theory Comput. 2019. PMID: 31507179 Free PMC article.
-
Mebendazole Mediates Proteasomal Degradation of GLI Transcription Factors in Acute Myeloid Leukemia.
Freisleben F, Modemann F, Muschhammer J, Stamm H, Brauneck F, Krispien A, Bokemeyer C, Kirschner KN, Wellbrock J, Fiedler W. Freisleben F, et al. Int J Mol Sci. 2021 Oct 1;22(19):10670. doi: 10.3390/ijms221910670. Int J Mol Sci. 2021. PMID: 34639011 Free PMC article.
References
-
- Bash PA; Field MJ; Davenport RC; Petsko GA; Ringe D; Karplus M, Computer Simulation and Analysis of the Reaction Pathway of Triosephosphate Isomerase. Biochemistry-Us 1991, 30 (24), 5826–5832. - PubMed
-
- Gao J, Catalysis by Enzyme Conformational Change as Illustrated by Orotidine 5′-Monophosphate Decarboxylase. Current Opinion in Structural Biology 2003, 13 (2), 184–192. - PubMed
-
- Ridder L; Mulholland AJ; Rietjens IMCM; Vervoort J, A Quantum Mechanical/Molecular Mechanical Study of the Hydroxylation of Phenol and Halogenated Derivatives by Phenol Hydroxylase. J. Am. Chem. Soc. 2000, 122 (36), 8728–8738.
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Other Literature Sources